Creating a confusion matrix between Baseline Model and Modified Model

While working on a problem statement, I came across the following, I would really appreciate if someone could explain to me what it is -

Train at least two models

  1. To establish a Baseline (using a pre-trained model)
  2. Your Modified algorithm (using the same pre-trained model)

Could someone explain to me how to create the two model and calculate a confusion matrix between them. Assuming I am working on Transfer Learning Tutorial as per the following link.

Any help will be appreciated.